What are your recommendations for squeezing the max requests per second (RPS) from a VC instance running on a DS_V2 Azure VM? (2 cores/7 GB RAM) with 100k products? If necessary, we can add external elasticsearch/redis cache.
Hi, we have 3 cache levels in our platform: 1 - level cache is within the VC platform API , 2 - level in storefront, 3 - level is ASP.NET full page caching. All this three levels together with indexed search and ajax product data actualization should provide high scale-ability and performance for you application on production.
Quantity of products does not should cause serious influence to resulting performance, thanks to using indexed search database structure optimization and caching.
Using Redis cache make sense to use only when you have multiple running storefront instances on web farm. For case with one storefront instance better use standard in-memory cache.
To confirm my words here are the results of http://demo.virtocommerce.com stress tests:
Used plan Standard 1 medium: 1 core and 1.75 GB